© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 26628
    Prev
    Next
    INDEX
    Explanations

    lists separated by commas and "and"

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    adamkarvonen/qwen3-32b-saes/saes_Qwen_Qwen3-32B_batch_top_k/resid_post_layer_32
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    евиÑĩ
    -0.10
    常å§Ķ
    -0.09
    èī¯å¿ĥ
    -0.09
    assed
    -0.09
    å½ĵä½ľ
    -0.08
    æ·¡æ·¡
    -0.08
    èģĮåĬ¡
    -0.08
    æ¸ħåĩī
    -0.08
    åIJĿ
    -0.08
    liner
    -0.08
    POSITIVE LOGITS
    thr
    0.10
    /or
    0.10
    ãģ¨ãģĦãģ£ãģ¦ãĤĤ
    0.10
    æĹı
    0.09
    ember
    0.09
    ex
    0.09
    ç¾İåĮĸ
    0.09
    fleet
    0.09
    åıįæŃ£
    0.09
    ree
    0.09
    Activations Density 0.171%

    No Known Activations